Materials and methods

In this study, to demonstrate the proof-of-principle, we used commercially available, fluorescently labeled avidin molecules conjugated to biotin on a polystyrene bead surface. Biotin and avidin were chosen because they have been widely used in biosensor development studies [19]. The experimental factors that contribute to the biotin-avidin conjugation are well-known and conjugation is straightforward to perform [19]. Therefore, we could focus on the engineering of the biosensor and determine strengths and limitations. Figure 3.1 shows the steps of our technique. First, we have conjugated avidin molecules to the complementary biotin molecules that are immobilized on the surfaces of commercially available polystyrene beads (Steps a–h).

We have then used the DEP force to concentrate the polystyrene beads in specific locations on an array of micro-interdigitated electrodes (IDE) (Step k). We will describe the steps of DEP based on the manipulation of the polystyrene beads below. Finally, we recorded a fluorescent image of the sample and developed a simple method to calculate the fluorescence intensity of the sample.

We then studied the variation of fluorescence with the concentrations of the avidin molecules.

Figure 3.1: Schematic representation of the steps of the biomarker detection technique. (a) Biomarker sample (1mL), (b) Biotinylated beads are added to the mixture, (c) Gentle vortexing to mix the contents in the sample, (d) Incubation on a shaker to conjugate biotin and avidin, (e) Conjugated biotin and avidin molecules, (f) Centrifuge to remove the free avidin molecules, (g) Sample after centrifugation, (h) Supernatant was removed and 50 µl of 0.01xPBS buffer was added, (i) Gentle vortexing to mixture sample, (j) 10 µl of sample was pipetted, and (k) Perform the detection and quantification.

The selectivity is dependent on the molecular type, conductivity of the medium and frequency of the external electric field applied in the sample [79]. Selective polarization produces a movement in the particle towards non-zero electric field gradient regions. For

example, polarized molecules move toward the highest (positive DEP) or lowest (negative DEP) electric field region, or it will stay stationary (zero force DEP) depending on the magnitude of polarization (positive, negative or zero) [79],[133],[134],[137]. From the DEP equation (1.4) the value of the CM factor is depends in the frequency-dependent dielectric properties of the

biomolecule and the surrounding medium, which plays an important role in determining the magnitude and direction of the DEP force [79],[93]. From the CM factor theoretical bound, it is

clear that the highest positive DEP force is greater than the highest negative DEP force [79],[133],[134],[137].

Figure 3.2: Design and fabrication of electrodes for DEP experiments. (a and b) Schematic view of the PIE electrodes. We have used the pearls shaped to produce large electric field s and field gradients needed to quickly detecting and quantifying biomolecules. (c) Scanning Electron Microcopy image of the fabricated PIE electrode. (d) Fluorescence image of the electrode when attractive DEP is applied and concentrated the beads in the highest electric field gradient regions, (d) calculated field gradient around one pearl of the PIE, (e) Fluorescence image of the electrode when beads are concentrated using negative DEP force.

In this study, we have used the combination of positive and negative DEP forces to attract and concentrate polystyrene beads in specific regions of the electrodes. First, we have used positive DEP (or attractive DEP) to concentrate the polystyrene beads in the electrode edges (Figure 3.2(d)). Moreover, attractive DEP will push free polystyrene beads toward the regions indicated by white arrows. We then changed the frequency of the electric field to produce

negative DEP (or repulsive DEP) on the polystyrene beads to repel them (Figure 3.2(d)) towards

regions indicated in orange arrow to concentrate them in a specific location between the

electrodes. Figure 3.2(f) shows the concentrated beads at specific locations on the electrode. The ability to manipulate polystyrene beads using DEP depends on the magnitude of the DEP force that is dependent on the electric field gradient produced by the electrodes. Therefore, the design of an electrode array that generates large electric field gradients in needed to produce larger DEP forces. We designed the electrodes using finite element modeling software and fabricated on a glass wafer. Interdigitated electrodes are commonly used for DEP experiments. Therefore, we have used our previous study as the basis [19] to develop an interdigitated electrode design capable of generating large DEP forces on polystyrene beads. Figure 3.2(a) shows the sketch of the electrodes with dimensions, and Figure 3.2 (b) shows a sketch of individual electrode pairs.

We call these electrodes as pearl-shaped interdigitated electrodes (PIE).

To calculate the expected electric field gradients, we have used COMSOL software and built a three-dimensional model using an AC/DC electric current (ec) physics module with frequency domain studies. We have assumed that an external voltage of 10 Vpeak-peak with 120 kHz frequency was applied to the electrode. Studies have used electric potentials with 120 kHz to generate large DEP forces on biomolecules, and this was sufficient to manipulate them [19].

Finally, we calculated the electric field gradients generated by our electrode design. Figure 3.2 (e) shows the variation of the electric field gradient near the electrodes (x, y, z=50 nm plane).

The calculated maximum and minimum electric field gradients were approximately 1x10¹⁴ V²/m³ and 10¹² V²/m³, respectively. We have compared these numbers with those for the standard interdigitated electrodes that were utilized in the literature, and our electrodes generate electric field gradients approximately 2–3 times higher than the traditional interdigitated electrodes [19],[79]. Therefore, these electrodes generate 2–3 times higher DEP forces on polystyrene

beads. A larger DEP force is expected to facilitate speedy manipulation of the beads and

eventually contribute to the detection speed of sensing. To fabricate the electrodes, we have first designed a photolithography mask in AutoCAD software and printed it on Mylar films using a dot-matrix printer (Fineline Imaging Inc, Colorado Springs, CO). The electrode was then

fabricated using standard photolithography processes followed by metal deposition and a lift-off process [54]. Figure 3.2(c) shows a scanning electron microscope image of fabricated electrode on glass wafer.

We purchased avidin molecules (Excitation: 495–500 nm, Emission: 514–521 nm) from Vector Laboratories (Burlingame, CA, USA), and biotinylated polystyrene beads (diameter=0.74 µm) were purchased from Spherotech Inc (Lake Forest, IL, USA). The selection of 0.74 μm as a bead size is based on our ability to manipulate the polystyrene beads using DEP force. In

particular, in previous studies, we determined that the polystyrene beads with a diameter ranging from 500 nm to 1 µm were easy to manipulate with DEP [54].

In experiments, we varied the molarity of the avidin molecules from µM to pM and measured the fluorescence of each concentration. Since the DEP force on a polystyrene bead is dependent on the number of biotin-avidin complexes on the surface, we have kept the number of biotin-avidin complexes on a bead constant for each avidin concentration by varying the number of polystyrene beads with concentrations of avidin molecules. The conjugation of biotin and avidin molecules was performed according to the manufacturer’s instructions. Briefly, to achieve 100% labeling of biotin molecules (approximately 10,000 molecules of biotin were on the surface of a bead) with avidin molecules, a 1:30 ratio of avidin molecules to biotinylated

polystyrene beads (1% w/v) was used. The sample (avidin molecules and biotinylated beads) was uniformly mixed with gentle vortexing for about 30 seconds. After that, the sample was kept on a

shaker (Ultra Rocker, BIO-RAD, Hercules, CA, USA) for 20 minutes at room temperature.

Finally, the sample was centrifuged for 20 minutes at 5000 rpm. The supernatant containing the unbound avidin molecules was removed, and 50 µl of 0.001 X PBS (σ=0.01 S/m) buffer was added to the tube [54]. We used low-conductivity buffer to avoid adverse effects when we use DEP force, such as electrolysis [54]. Studies have reported that biotin-avidin duplexes were stable in this low-conductivity buffer. Finally, the sample was uniformly mixed by gently vortexing for about 30 seconds. We have taken many precautions during this assay to make sure that we will have an intact sample of biotin-avidin duplexes. For example, we have covered the centrifugation tube with aluminum foil to avoid bleaching the avidin molecules by exposure to light.

To prepare the electrodes for experiments, we have cleaned the electrodes using a 75%

ethanol solution, washed them in DI water, and then dried them using pressurized air [19]. The clean electrode was placed firmly on the electrode holder using commercially available adhesive tape, and the electrode holder with the electrodes was mounted on a low-power fluorescence microscope (Omano, OMFL600, Roanoke, VA, USA). Next, the electrical connections to the electrodes were made by connecting them to a function generator (Tektronix, AFG 3021B, Beaverton, OR, USA). We then determined the frequencies needed to generate positive and negative DEP forces on the beads. Studies have indicated that polystyrene beads experience positive DEP forces at lower frequencies (<500 kHz) (6, 16-18). Therefore, we varied the frequency below 500 kHz and determined the most suitable frequency (10 Vp-p and 10 kHz) needed to generate the largest DEP force on the polystyrene beads with conjugated biotin-avidin molecules. Similarly, to find the appropriate negative DEP frequency (repulsive DEP), we varied the frequency from 500 kHz–5 MHz. The largest repulsive DEP force was generated at 3 MHz.

These attractive and repulsive DEP forces were calculated by recording videos when beads were moving with positive or negative DEP frequencies. We then calculated the velocities of beads because velocities are proportional to the DEP force on the beads. These frequency values of the positive and negative DEP forces are dependent on various factors, such as the conductivity of the buffer, the type of molecules that we have on the surface of the polystyrene beads, and the diameter of the beads [79],[133],[134],[137].

Figure 3.3: Results from the dielectrophoretic based detection and quantification of Avidin molecules. (a) Variation of fluorescence with molarity of the Avidin molecules, (b and c)

Standard curve to be used in the finding molarity of unknown sample. (b and c) shows the

variation of fluorescence with number of pixels for avidin molecules that were suspended in PBS buffer and serum respectively.

After figuring out the frequency values, we have performed avidin detection experiments.

The first experiment was performed by spiking in avidin molecules of varying concentrations (from μM to pM) in 1X PBS buffer (positive control). The second experiment was performed by

spiking avidin molecules (μM to pM concentrations) into diluted serum samples (1:99=serum:

DI water).

After conjugating biotin and avidin, we suspended the beads in 0.001X PBS buffer and loaded the sample on the electrodes to measure the fluorescence. First, we have applied the positive DEP force for about 15 seconds to bring biotin-avidin-labeled polystyrene bead electrode edges from all three x, y, and z directions. Then negative DEP was applied to

concentrate beads in the regions where there was the lowest electric field gradient. Finally, we recorded a fluorescent image of the sample [79],[133],[134],[137] . Each experiment was repeated to assess reproducibility.

We used ImageJ software to analyze the fluorescent images (https://imagej.nih.gov/ij/).

First, we determined a set of parameters to enhance the brightness, sharpness, and contrast of the images. We used these parameters to process all the images. We then extracted the fluorescent intensities of each pixel from each modified image by using a simple, custom-built software program. We then plotted the extracted fluorescence intensity of each pixel of the image as histograms. Figure 3.3(a) shows the variation in fluorescence intensities with respect to the concentration of the avidin molecules. As indicated in Figure 3.3(a), there was a significant enhancement in the fluorescence when DEP force was used to concentrate polystyrene beads with conjugated avidin and biotin molecules (see the bar charts for 0.15 M with and without an applied DEP force). We then calculated the total fluorescence of each image for each

concentration. We used the fluorescence intensities that were above 70, and the total fluorescence intensity (𝛺) was defined as follows:

𝛺 = ∑²⁵⁶_𝑖=70𝐼_𝑖. 𝑛_𝑖 (3.1)

where ni is the number of pixels corresponding the Ii. We then plotted the variation of Ω with the molarity of avidin molecules (Figures 3.3(b–c)). Figure 3.3(b) shows the detection of the avidin molecules that were spiked into PBS buffer, and Figure 12c shows the avidin detection data from the diluted serum samples. Note that the fluorescence intensity varies with the molarity of avidin in both samples. We then selected a concentration that can be expected in early state disease progression (0.15 M) for comparison, Figure 3.3(b) shows the comparison of fluorescence intensity with and without DEP-based clustering. Note that the concentration of polystyrene beads using DEP force has increased the total fluorescence of the sample by 100-fold when compared to the fluorescence of the sample that did not experience DEP force. The 100-fold improvement is significant because it will improve the detection limit by at least 100 times.

Figure 3.3(c) shows the detection of various molarities of avidin from diluted serum samples.

The smallest molarity that we could detect was 1.5 pM. In comparison, the smallest molarity that ELISA can detect is about 40 nM [138].